THE page reports on six active exploits, detailing vulnerabilities such as Ajax.NET's deserialization issue and Microsoft's SQL Server's remote code execution risk. Additionally, it introduces Google's Gemini 3.5 Transcribe, a speech-to-text model that enhances transcription accuracy and reduces latency, capable of handling over 85 languages and supporting up to three speakers during transcription.
The model offers APIs for real-time and pre-recorded applications, achieving lower word error rates and faster transcription speeds than previous models. It is currently available in public preview.